From 6ef4fb387d50fa8f3bffdffc868b57e981cdd709 Mon Sep 17 00:00:00 2001 From: Mark Rutland Date: Tue, 3 Jan 2017 14:27:26 +0000 Subject: arm64: mm: fix show_pte KERN_CONT fallout Recent changes made KERN_CONT mandatory for continued lines. In the absence of KERN_CONT, a newline may be implicit inserted by the core printk code. In show_pte, we (erroneously) use printk without KERN_CONT for continued prints, resulting in output being split across a number of lines, and not matching the intended output, e.g. [ff000000000000] *pgd=00000009f511b003 , *pud=00000009f4a80003 , *pmd=0000000000000000 Fix this by using pr_cont() for all the continuations. Acked-by: Will Deacon Signed-off-by: Mark Rutland Signed-off-by: Catalin Marinas --- arch/arm64/mm/fault.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/arch/arm64/mm/fault.c b/arch/arm64/mm/fault.c index a78a5c401806..156169c6981b 100644 --- a/arch/arm64/mm/fault.c +++ b/arch/arm64/mm/fault.c @@ -88,21 +88,21 @@ void show_pte(struct mm_struct *mm, unsigned long addr) break; pud = pud_offset(pgd, addr); - printk(", *pud=%016llx", pud_val(*pud)); + pr_cont(", *pud=%016llx", pud_val(*pud)); if (pud_none(*pud) || pud_bad(*pud)) break; pmd = pmd_offset(pud, addr); - printk(", *pmd=%016llx", pmd_val(*pmd)); + pr_cont(", *pmd=%016llx", pmd_val(*pmd)); if (pmd_none(*pmd) || pmd_bad(*pmd)) break; pte = pte_offset_map(pmd, addr); - printk(", *pte=%016llx", pte_val(*pte)); + pr_cont(", *pte=%016llx", pte_val(*pte)); pte_unmap(pte); } while(0); - printk("\n"); + pr_cont("\n"); } #ifdef CONFIG_ARM64_HW_AFDBM -- cgit v1.2.3 From 9d84fb27fa135c99c9fe3de33628774a336a70a8 Mon Sep 17 00:00:00 2001 From: Mark Rutland Date: Tue, 3 Jan 2017 18:27:01 +0000 Subject: arm64: restore get_current() optimisation Commit c02433dd6de32f04 ("arm64: split thread_info from task stack") inverted the relationship between get_current() and current_thread_info(), with sp_el0 now holding the current task_struct rather than the current thead_info. The new implementation of get_current() prevents the compiler from being able to optimize repeated calls to either, resulting in a noticeable penalty in some microbenchmarks. This patch restores the previous optimisation by implementing get_current() in the same way as our old current_thread_info(), using a non-volatile asm statement. Acked-by: Will Deacon Signed-off-by: Mark Rutland Reported-by: Davidlohr Bueso Signed-off-by: Catalin Marinas --- arch/arm64/include/asm/current.h | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/arch/arm64/include/asm/current.h b/arch/arm64/include/asm/current.h index f2bcbe2d9889..86c404171305 100644 --- a/arch/arm64/include/asm/current.h +++ b/arch/arm64/include/asm/current.h @@ -9,9 +9,17 @@ struct task_struct; +/* + * We don't use read_sysreg() as we want the compiler to cache the value where + * possible. + */ static __always_inline struct task_struct *get_current(void) { - return (struct task_struct *)read_sysreg(sp_el0); + unsigned long sp_el0; + + asm ("mrs %0, sp_el0" : "=r" (sp_el0)); + + return (struct task_struct *)sp_el0; } #define current get_current() -- cgit v1.2.3